第14期

从原始医学影像到虚拟临床试验：面向人类心脏的自动化框架

From Raw Medical Imaging to Virtual Clinical Trials: An Automated Framework for the Human Heart

报告人：Roberto Verzicco   欧洲力学学会主席、意大利格兰萨索科学研究所教授

主持人：何国威 院士   中国力学学会理事长、中国科学院力学研究所研究员

时   间：2026年4月10日（周五）上午10:00

地   点：中国科学院力学研究所5号楼206会议室

参会方式：线下参与，需扫描下方二维码报名；线上观众可通过下方直播卡片，预约观看。

▲如线下参与讲座，请扫码提交信息

报告摘要：

Because of the variability among human beings, the efficacy of novel treatments can only be assessed through statistical analysis on large cohorts of patients. However, finding a sufficient number of cases for rare or multiple pathologies could be an impossible task thus preventing these subjects from accessing innovative treatments.

Producing synthetic data by high–fidelity computer models might be a viable strategy to mitigate the above limitations and cardiovascular disorders are among the most promising fields of digital medicine.

In fact, when supplied with suitable input parameters, these computer models can surrogate real patients with ‘on demand’ features thus allowing cost effective clinical trials for those pathologies in which enrolling real subjects would be unpractical.

Reliable cardiovascular simulations, however, entail modelling and simulating the various heart systems, such as hemodynamics, electrophysiology and active tissue elastomechanics including their interconnected physics.

Another, often underestimated, bottleneck is the availability of a sufficient number of computational–ready heart configurations, both for physiological and pathological cases, to perform significant statistical analysis.

In this seminar, we show some of the latest activities of the Fluid Dynamics Group at GSSI on cardiovascular flows. We have integrated computational methods coming from different disciplines to build an automated framework capable of producing datasets of diagnostic quantities for human hearts starting from their raw medical images (CT–or MRI–scans).

In order to achieve this goal, we have combined standard techniques (direct numerical simulation and fluid/structure interaction) with more recent ones (immersed boundary methods and PCA–analysis) and the whole model has been implemented using the latest GPU–acceleration technologies. The resulting tool duplicates the full cardiac dynamics with unprecedented precision and time–to–solution compatible with clinical practice.

This allows extensive simulation campaigns to study the response of synthetic cohorts of patients to cardiovascular disorders, novel prosthetic devices or surgical procedures. We show the results obtained for selected pathologies, induced by altering the physiological parameters. The in–silico results closely match those obtained in clinical practice, confirming the reliability of the framework.

Figure 1. Perspective views, for the healthy configuration, of instantaneous distributions of activation potential (a) and (d), internal tissue stress (b) and (e) and blood velocity magnitude on a plane cutting the left ventricle (c) and (f). a), b) and c) are for late diastole, d) e) and f) for peak systole.

报告人简介：

Roberto Verzicco is professor of Fluid Dynamics at Gran Sasso Science Institute in L'Aquila, Italy. Since 2010 he is also part-time Professor at the University of Twente. He received his M.S. in Aeronautic Engineering (1991) and Ph.D. in Aerospace Engineering (1994) both from University La Sapienza in Rome. His research, in the fields of experimental and numerical fluid dynamics, includes direct numerical simulation of turbulent flows, convective turbulence and bio-fluid mechanics. Since 1993 he has published more than 300 papers in international journals with 3 solicited reviews. He has published a monographic book on Immersed Boundary Methods with Cambridge University Press. He has organized several international conferences, workshops and colloquia and in 2012 he chaired the 9th European Fluid Mechanics Conference in Rome. Since 2012 he is Associate Editor of the Journal of Fluid Mechanics and since 2025 he is the President of Euromech. In November 2005 he received the Frenkiel Award 2005 from the American Physical Society for his studies on natural convection and in 2022 the Corrsin Award for the immersed boundary method. In 2012 he was elected EUROMECH Fellow and in 2013 APS Fellow.

本场讲座将于线上线下同步进行，我们诚挚邀请各位学者、研究人员及学生参加此次讲座，与大师面对面，共同探索从原始医学影像到虚拟临床试验的人类心脏自动化框架！

如报名后确认当天不能到场，请提前告知。

联系人：张烨文

电话：010-82543905

邮箱：zhangyewen@cstam.org.cn